Upon completing her second degree, Anisa* expressed her confidence in her abilities to find employment, citing her postgraduate studies, double major undergraduate degree, and years of juggling two volunteer positions and a part-time job as contributing factors.
She kept a spreadsheet track of the results of every application she submitted for entry-level or junior positions in her industry. “Applying for jobs became a full-time job before she knew it,” she said. She applied for and was turned down for 350 jobs overall before 18 months later, she was hired. She also thinks that a major contributing factor in this is AI, particularly as it relates to its use in screening applications.
Anisa remembers uploading resumes in addition to filling out digital forms, and she says that the rise of third-party, AI-run digital online forms is a major source of frustration for many job seekers.
The number of personalized interview processes she have had where she could assign a face is so small that it can be counted on one hand.
Even though it’s relatively new, AI is widely used in the recruitment industry in Australia. Researchers estimate that over 250 commercial AI recruitment tools are currently being used in Australia. Hiring process components that were previously the purview of humans, such as resume screening and preliminary interviews, are increasingly being outsourced to AI. One in three Australian organisations said as much as a year ago that they had used AI tools for hiring, despite the fact that some researchers have classified this as a “high-risk” activity.
Increasing the use of AI in the hiring industry, according to John Shields, professor of human resource management and organizational studies at the University of Sydney Business School, is not generally a bad thing.
According to him, the key to minimizing the problems and expenses associated with a poor hiring decision is to swiftly and efficiently manage enormous volumes of candidate data and establish a solid alignment between the individual and the organization, or the individual and the job.
Employers are using AI because the conventional hiring process—application, interview, work test, referee contact—is “outdated, onerous and time-consuming” and does not always produce the best candidate for a position, according to Sue Williamson, associate professor of human resources management at UNSW.
According to her, managers are becoming more time-constrained and are seeking for methods to streamline the hiring process. Employers might be using AI to help expedite administrative procedures or to help screen applicants by, say, “assessing” resumes or giving work tests to prospective employees. AI is also capable of evaluating video resumes, which are also growing in popularity.
However, it raises concerns about merit and fairness, which is a sore point for many job seekers. AI-assessed job applications reinforce prejudices against women and cultural minorities, according to recent research. After overturning 11 promotional decisions made by Services Australia in the 2021–2022 fiscal year due to the use of AI–assisted and automated selection techniques without human review, the Merit Protection Commissioner in Australia released new guidelines for public sector employers. This resulted in “meritorious applicants missing out on promotions.”
The most well-known study on biases in hiring and selection comes from Amazon, where programmers flooded the AI (resume screening tool) with male resumes, according to Williamson. After then, the AI discriminated against women’s resumes by favouring those from men. Research has additionally demonstrated that, in comparison to candidates with western names, candidates with non-Western names—such as Arabic and Asian names—also receive far fewer invitations to interviews.
Job search depression’
The lack of clarity surrounding the hiring process, and more especially, how candidates will be evaluated, is increasing candidate anxiety. According to Williamson, there are no laws in Australia that specifically mandate that candidates be informed about AI screening procedures.
She says that this could be problematic because it could result in indirect discrimination against certain groups of job seekers. For example, requiring applicants to complete online tasks during the recruitment process could be detrimental to those who lack manual dexterity.
Anisa lists the time spent on different tasks, the absence of feedback on applications, and the psychological and physical costs of going through such a repetitive process—which frequently resulted in what she refers to as “brutal” rejections—as sources of annoyance and bitterness.
She claims that the entire process—from the searching and application stages to the multiple interview stages and [consecutive] rejections to the hundreds of times she was just ghosted—was exhausting, demoralizing, and stressful.
Since a job search typically takes five months, this resentment can also result in “job search depression,” which experts warn can cause procrastination, prolonged anxiety, and maladaptive coping mechanisms like increased drinking, sleeping, or disordered eating.
After 15 years in the industry, Lisa* applied for the same position at a different company and discovered that AI had recorded her interview. A question would be posed to her, she would have sixty seconds to answer it, and her response would be recorded.
Before starting his current position, data analyst Frank* recalls how “nerve-wracking” it was to respond to a 100-question “psych-test.” He has since stopped doing any jobs that employers initially ask candidates to complete.
The initial step should always be to conduct a face-to-face interview with candidates to learn more about them. Anisa concurs. According to her, “timed ‘tests’ are also used for many screening interviews.” “I completed one [for] a multinational tech company that required both written and… video responses. I didn’t meet or talk to anyone—not even over the phone—before receiving a generic rejection email, and it felt incredibly impersonal. In a different role, we had to solve maths problems, calculate service costs after inflation, and answer an open-ended question about how much money you would need to live comfortably for the rest of your life without working.
Biases and unfairness
However, Shields asserts that assessments of cognitive ability are not novel.
We have to be careful not to claim that these things are the phenomenon of the age of the robot, he says, because no [candidate] can succeed unless they are quite literate or quite numerate.
Use of candidate-generated online videos is novel; candidates are ranked according to characteristics or attributes that the AI platform is analogously programmed to determine as desirable or undesirable based on the responses to these videos, which are subsequently analyzed by AI to create candidate profiles. Prior to a candidate having a face-to-face or panel interview, candidate-recorded interviews will play a bigger and bigger role in the screening process.
Anisa has spoken of the brutality of the rejections she has faced while looking for work; Lisa* believes that the process ought to be two-sided and that right now things are crazy.
According to her, the purpose of the initial interviews is to learn more about the company and how you envision yourself working there. For [a] business to believe that an interview should be one-sided feels conceited.
She says that the first interviews are essential to learning more about the company and how you envision yourself working there. The idea that an interview should be one-sided seems conceited on the part of [a] company.
However, Shields notes that when utilized appropriately, technology can benefit both employers and employees by helping them match the best candidate with the job.
He says that if the AI platform’s recommendation-determining programming is biased, then garbage in, garbage out. Ensuring that analogous rules for AI platforms are literally written to iron out biases and unfairness requires careful attention to detail.
Thus, what can a job seeker do?
When applying for jobs, Shields advises applicants to follow employer expectations and be clear about their strengths. She also stresses the importance of a cover letter that is forensically focused on the selection criteria and that resumes be concise, precise, and accurate.
Additionally, he cautions job seekers about leaving a trace online. He advises against having anything on your media profile that could cast doubt on your moral character, behaviour, or political beliefs if you’re serious about landing the job.
